Well, thanks Sudha,
First of all, I'm sorry that I missed to state that I lyophilized the protein after
dialysis. The protein is a well behaviored one. Almost complete renaturation is
expected after denatured in GdmHCl. So there shouldn't be problem concerning the
concentration.
In my case, the effective concentration should be 4mg dimer/500uL for doubly labelled
protein, 2mg/500uL asymmentricaly labelled dimer for the mixed sample. This gives a
ratio of 2. Concerning the NOE transfer pathway, a factor of 2 should be considered
for the doubly labelled sample. Thus the comparison of s/n of interfacial NOE of
doubly labelled sample and the mixed sample should be:
caculated s/n comparison
s/n(double) 2 sqrt(32)
----------- =-- x 2 x ------------=1/2
s/n(mixed) 1 sqrt(2048)
Surely, as pointed out by Weixing Zhang of stjude.org, the NOEs I got for the mixed
sample were interfacial only while those for the double labelled sample include
intramolecular ones, which actually prevail compared with that of the interfacial
NOEs, thus low s/n is to be expected. What to MY PUZZLE is that should such low an
s/n be normal:
measured s/n comparison
s/n(double) 31
----------- = ---
s/n(mixed) 2
This leads to an interfacial NOE only ~3% of the intramolecular NOE.
If the above calculation and deduction are correct, then there comes the problem of
how to discriminate true interfacial NOEs and intramolecular NOEs originate from
natural abundant 13C, 15N:
A 15N singly labelled protein holds 1.1% natural abundant 13C in the molecular, this
will also give NOE in a 13C,15N HMQC-NOESY-HSQC experiments.
Experiment on a 15N labelled protein confirms this:
I did observe similiar NOE peaks in a ns=2048 1D version of 4D HMQC-NOESY-HSQC in the
15N singly labelled sample, compared with that for the mixed sample. Though not
identical, the s/n's are similiar.
I hope I was wrong somewhere.
Regards,
Yu-Feng
Sudha Veeraraghavan wrote:
> 1. When a dimer is refolded with equal amounts of labeled and unlabeled protein (or
> differently labeled proteins), one gets 25% fully labeled dimers, 25% fully
> unlabeled dimers, and 50% asymmetrically labeled dimers. So, in your case, the
> signals you would expect from the doubly labeled protein should be from 25% of
> total protein (the fraction that has one monomer with 15N and the other monomer
> with 13C). So, the S/N will not be as high as you might otherwise expect.
> 2. Do you know if you have really have 8mg/500 uL total protein concentration? How
> did you verify the concentration? I ask this because many proteins tend to crap out
> to some extent or the other during refolding.
>
> To make sure loss of protein during refolding is not the problem, one thing to try
> - if you haven't already- is to do the refolding on unlabeled molecules. Check
> soluble protein concentration and NMR spectra before and after refolding (simple 1D
> NOE) to prove to yourself the % recovery after folding.
>
> Hope this helps,
> Sudha Veeraraghavan
>
-----Original Message-----
> From: Yu-Feng Tong [mailto:nmradm_at_nlbnmr.ibp.ac.cn]
> Sent: Tuesday, March 05, 2002 6:44 PM
> To: ammrl_at_chemnmr.colorado.edu
> Subject: Question about the sensitivity of HMQC-NOESY-HSQC experiment
Hello all,
I'd like to know anyone had the experience on carrying out 4D
13C,15N-separated HMQC-NOESY-HSQC experiment and/or the experience on
solving the structure of dimer. Here is my puzzle:
I'm trying to determine the interfacial NOE of a dimer ( 97 a.a.
monomer, m.w. ~11,000). I have prepared an 8mg/500uL equal molar ratio
15N/13C asymmetric labelled sample (4mg 15N labelled, 4mg 13C labelled,
mixed and denatured by 6M GdmHCl, incubated at 4 degree overnight, then
dialysized to renature the protein), then I carry out the 1D version of
a 4D 13C,15N HMQC-NOESY-HSQC (Muhandiram, Xu, Kay: J. Biomol. NMR 3,
463-70 (1993)), i.e. acquire the first fid of a 4D experiment.
To my surprise, the s/n ratio is very low even I accumulate 2048
transients. Actually the s/n is 2 after 2k scans. For a comparison, I
found the s/n of a 15N,13C double labelled sample (8mg/500uL too,
identical acquisition parameters) to be 31 after 32 scans for a 1D
version HMQC-NOESY-HSQC.
I have suspected the protein to be a monomer, but size exclusion
chromatography data show the protein to be larger than a 150 a.a.
protein. And I also found Tracy M. Handel had used a ns=2048 for 2D
version of the 4D experiment to identify interfacial NOE for MCP-1
protein, (yet ns=64 for a 3D version)
Any hint is greatly appreciated.
Yu-Feng
--
Yu-Feng Tong||童宇峰
NMR Group, Dept. Enzymology
Inst. Biophysics, Academia Sinica
15 Datun Road, Beijing, P.R.China, 100101
_______________________________________________________________________________
Life is an everlasting game of Weiqi(igo, baduk).
Received on Wed Mar 06 2002 - 09:41:33 MST